Gavyn Davies has launched a very interesting debate on whether or not the upcoming completion of QE2 will result in a surge in Treasury yields, as a key driver of recent bond demand disappears from the market. One view is that a decline in the flow demand for bonds will, all else equal, cut the price and force up yields, effectively tightening monetary conditions even before the Fed raises its targets for short-term interest rates. Davies takes the view that a rise in yields is not inevitable, citing portfolio balance theory: Given that the Fed has absorbed a share of the overall stock of Treasuries, and will not put them back on the market for some time, the risk of future flight from bonds will decline, and, as a result, investors will accept a lower yield on Treasuries because the Fed’s intervention has made them an even safer asset class in terms of future capital value. This equates to investors being willing to pay a higher price for the flow of new Treasuries going forward, effectively expanding the demand curve in that market and keeping prices high and yields low despite the loss of the Fed as a net buyer of bonds. In effect, private demand for Treasuries emerges to replace Fed demand.
I tend to side with the view that the Fed’s exit will not induce a surge in yields. Here are some thoughts on what might prop up the flow demand for Treasuries in order to stabilize prices, and therefore yields, even after the end of QE2.
(1) Re-investment of maturing funds.
A policy of maintaining the overall size of the Fed’s balance sheet requires re-investing the funds accrued from assets that mature, as seen in the summer of 2010 when the central bank cashed in some of the mortgage backed securities acquired in the early stages of QE. The Treasuries purchased by the Fed are long-term assets and will not necessitate re-investment decisions. However, a far as I can see, the Fed still has plenty of assets on its books due for redemption in the next couple of years, and re-investment decisions then arise. One possibility is that at least some component of the maturing funds will be used to acquire Treasuries, potentially sustaining the Fed’s flow demand for this asset class beyond the formal QE2 endpoint. In that case, QE could continue to prop up Treasury prices and contain yields without a commitment to QE3 and beyond. Of course, it may be the case that the Fed chooses zero re-investment of funds, but in the event that it acts to prevent natural attrition of its balance sheet, Treasury purchases seem a more likely prospect than do other kinds of investment.
(2) Investor demand for Treasuries.
The debate about exactly which markets benefited from QE will continue for a long time. Things that we seemed to have learned so far are that central bank liquidity injections expanded the narrow money supply much faster than broad money measures (there was no boom in bank lending on the back of QE), but there are clear signs that QE boosted the shadow banking system in the form of equity markets and corporate bond markets. But there seems no way of saying that the entire stock of Fed liabilities (cash reserves) generated as the flipside of asset purchases has been invested in full. It seems quite possible that major institutional investors still hold more cash in their portfolios than they would have done absent QE, and that eventually these funds will be invested in a way that benefits Treasuries. This is basically suggesting that in addition to portfolio balance effects generating some private flow demand for Treasuries, there could be a similar effect on the flow demand because QE has left portfolios out of equilibrium and there is some adjustment still to happen.
(3) The policy rate outlook
As others have noted, the trajectory for Treasury and related yields will depend on the expected path of the Fed’s target for the federal funds rate, currently set to remain at historic lows for an extended period, and the deposit rate that the Fed elects to pay for excess reserves held at the central bank. In the event of a rise in policy rates, the excess liquidity discussed in (2) would be more likely held on deposit at the Fed, in which case the flow demand for Treasuries would be stymied and yields would be expected to rise (though for some thoughts on the limits to the Fed’s ability to drain excess reserves from the system in this way, see this piece). A rise in policy rates would have similar effects to the decision not to re-invest maturing funds. Despite the improving performance of the US labor market, I think both of these moves are unlikely in the short-term, given that the US needs to achieve some of the increase in inflation expectations that Krugman and others have argued is necessary to improve growth prospects.
A final thought: The UK comparison tells us that after the end of QE, yields did not rise dramatically, which is an important point to bear in mind. In the UK, I think the situation in terms of (2) and (3) is very similar to what we observe as the US moves towards the end of the current phase of QE. However, (1) was much less likely to have applied in the UK when QE ended in 2010, given that QE at the Bank of England was dominated by gilt purchases. This suggests a surge in yields in the US is less likely than it was in the UK when it confronted an official pause in QE. Of course, in all of this nobody can be proved right or wrong given the other factors that will impinge on bond prices going forward – inflation expectations, the changing supply of bonds driven by government financing requirements, and the unfolding political situation to name just a few.